ICS 321 Spring 2011 Introduction to Database Systems Asst. Prof. Lipyeow Lim Information & Computer Science Department University of Hawaii at Manoa 1/12/20111Lipyeow Lim -- University of Hawaii at Manoa

Group Discussion How would you implement an internet store without using any database software ? – List all the questions, challenges or difficulties you encounter as you discuss the high level steps. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa2

Group Discussion Results 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa3

Data, Database, DBMS A database : a collection of related data. Represents some aspect of the real world (aka universe of discourse). Logically coherent collection of data Designed and built for specific purpose Data are known facts that can be recorded and that have implicit meaning. A data model is a collection of concepts for describing data. A schema is a description of a particular collection of data, using the a given data model. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa4

DBMS A database management system (DBMS) is a collection of programs that enables users to – Create new DBs and specify the structure using data definition language (DDL) – Query data using a query language or data manipulation language (DML) – Store very large amounts of data – Support durability in the face of failures, errors, misuse – Control concurrent access to data from many users 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa5

Types of Databases On-line Transaction Processing (OLTP) Banking Airline reservations Corporate records On-line Analytical Processing (OLAP) Data warehouses, data marts Business intelligence (BI) Specialized databases Multimedia XML Geographical Information Systems (GIS) Real-time databases (telecom industry) Special Applications Customer Relationship Management (CRM) Enterprise Resource Planning (ERP) Hosted DB Services Amazon, Salesforce 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa6

A Bit of History 1970 Edgar F Codd (aka “Ted”) invented the relational model in the seminal paper “A Relational Model of Data for Large Shared Data Banks”  Main concept: relation = a table with rows and columns.  Every relation has a schema, which describes the columns. Prior 1970, no standard data model. Network model used by Codasyl Hierarchical model used by IMS After 1970, IBM built System R as proof-of-concept for relational model and used SQL as the query language. SQL eventually became a standard. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa7

Files vs DBMS Swapping data between memory and files Difficult to add records to files Security & access control Do optimization manually Good for small data/files Run out of pointers (32bit) Code your own search algorithm Search on different fields is difficult Must protect data from inconsistency due to concurrency Fault tolerance – crash recovery 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa8

Why use a DBMS ? Large datasets Concurrency/ multi- user Crash recovery Declarative query language No need to figure out what low level data structure Data independence and efficient access. Reduced application development time. Data integrity and security. Uniform data administration. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa9

DBMS Components 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa10 Storage Storage Manager Buffer Manager Index/file/record Manager Execution Engine Query Compiler Transaction Manager Logging & Recovery DDL compiler Concurrency Control Lock Table Buffers User/Application Database Administrator

Transaction: An Execution of a DB Program A transaction an atomic sequence of database actions (reads/writes). Each transaction, executed completely, must leave the DB in a consistent state if DB is consistent when the transaction begins.  Users can specify some simple integrity constraints on the data, and the DBMS will enforce these constraints.  Beyond this, the DBMS does not really understand the semantics of the data. (e.g., it does not understand how the interest on a bank account is computed).  Thus, ensuring that a transaction (run alone) preserves consistency is ultimately the user’s responsibility! 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa11

Concurrency Control Concurrent execution of user programs is essential for good DBMS performance.  Because disk accesses are frequent, and relatively slow, it is important to keep the cpu humming by working on several user programs concurrently. Interleaving actions of different user programs can lead to inconsistency: e.g., check is cleared while account balance is being computed. DBMS ensures such problems don’t arise: users can pretend they are using a single-user system. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa12

ACID Properties Atomicity : all-or-nothing execution of transactions Consistency: constraints on data elements is preserved Isolation: each transaction executes as if no other transaction is executing concurrently Durability: effect of an executed transaction must never be lost 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa13

Ensuring Isolation Scheduling concurrent transactions DBMS ensures that execution of {T1,..., Tn} is equivalent to some serial execution T1’... Tn’.  Before reading/writing an object, a transaction requests a lock on the object, and waits till the DBMS gives it the lock. All locks are released at the end of the transaction. (Strict 2PL locking protocol.)  Idea: If an action of Ti (say, writing X) affects Tj (which perhaps reads X), one of them, say Ti, will obtain the lock on X first and Tj is forced to wait until Ti completes; this effectively orders the transactions.  What if Tj already has a lock on Y and Ti later requests a lock on Y? (Deadlock!) Ti or Tj is aborted and restarted! 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa14

Ensuring Atomicity DBMS ensures atomicity (all-or-nothing property) even if system crashes in the middle of a Xact. Idea: Keep a log (history) of all actions carried out by the DBMS while executing a set of Xacts:  Before a change is made to the database, the corresponding log entry is forced to a safe location. (WAL protocol; OS support for this is often inadequate.)  After a crash, the effects of partially executed transactions are undone using the log. (Thanks to WAL, if log entry wasn’t saved before the crash, corresponding change was not applied to database!) 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa15

The Log The following actions are recorded in the log:  Ti writes an object: The old value and the new value.  Log record must go to disk before the changed page!  Ti commits/aborts: A log record indicating this action. Log records chained together by Xact id → easy to undo a specific Xact (e.g., to resolve a deadlock). Log is often duplexed and archived on “stable” storage. All log related activities (in fact, all CC related activities such as lock/unlock, dealing with deadlocks etc.) are handled transparently by DBMS. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa16

Summary Definitions of data, databases, data models, schema When to use or not use a DBMS DBMS major components Transactions and concurrency ACID properties of transactions Techniques for ensuring ACID properties in DBMSs. 1/12/2011Lipyeow Lim -- University of Hawaii at Manoa17